EP1520384B1 - Verfahren und vorrichtung zur gleichzeitigen schätzung des frequenzversatzes und modulationsindex - Google Patents

Verfahren und vorrichtung zur gleichzeitigen schätzung des frequenzversatzes und modulationsindex Download PDF

Info

Publication number
EP1520384B1
EP1520384B1 EP03761510A EP03761510A EP1520384B1 EP 1520384 B1 EP1520384 B1 EP 1520384B1 EP 03761510 A EP03761510 A EP 03761510A EP 03761510 A EP03761510 A EP 03761510A EP 1520384 B1 EP1520384 B1 EP 1520384B1
Authority
EP
European Patent Office
Prior art keywords
estimator
value
modulation index
filters
frequency offset
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
EP03761510A
Other languages
English (en)
French (fr)
Other versions
EP1520384A1 (de
Inventor
Gerrit Smit
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Telefonaktiebolaget LM Ericsson AB
Original Assignee
Telefonaktiebolaget LM Ericsson AB
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Telefonaktiebolaget LM Ericsson AB filed Critical Telefonaktiebolaget LM Ericsson AB
Publication of EP1520384A1 publication Critical patent/EP1520384A1/de
Application granted granted Critical
Publication of EP1520384B1 publication Critical patent/EP1520384B1/de
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/18—Phase-modulated carrier systems, i.e. using phase-shift keying
    • H04L27/22—Demodulator circuits; Receiver circuits
    • H04L27/233—Demodulator circuits; Receiver circuits using non-coherent demodulation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/0014—Carrier regulation

Definitions

  • the present invention relates generally to the field of radio receivers that utilize continuous phase modulation (CPM) and, more particularly, to a method of and system for estimating a modulation index and a carrier frequency offset of a CPM signal.
  • CPM continuous phase modulation
  • Wireless technologies such as terrestrial and satellite mobile communications and/or BLUETOOTH systems, may use continuous-phase-modulated (CPM) signals to transmit data.
  • CPM continuous-phase-modulated
  • Binary CPM or M-ary CPM may be employed for the wireless transmission of data packets.
  • the modulation index may need to be known in some receiver architectures.
  • receiver architectures could be employed that require knowledge of the value of the modulation index of the transmitted signal. Due to the use of independent frequency generating circuits in the transmitting and receiving devices, a carrier frequency offset is typically generated. In order to achieve optimal performance, the carrier frequency offset should be compensated for as much as possible. Therefore, there is a need for a method of and system for estimating the modulation index and the carrier frequency offset of a CPM signal.
  • U.S. Patent No. 6,389,040 discloses a method and apparatus for generating a frequency offset.
  • a received signal is mixed with a locally generated frequency corresponding to a frequency offset to generate a mixed signal.
  • a calculation is made on the mixed signal in which channel outputs of the same state are combined and accumulated. Then a summation is made over all possible states of the combined and accumulated channel outputs to yield a so-called metric calculation value for that mixed signal.
  • the metric calculation is then repeated for a plurality of different locally generated frequencies corresponding on a one-to-one basis with a plurality of frequency offsets.
  • the frequency offset corresponding to the largest metric calculation value is selected as the desired frequency offset estimate.
  • U.S. Patent No. 6,393,073 discloses a method of frequency offset estimation and correction for adaptive antennas.
  • the method includes receiving, in a processor, samples of a data set having a training data sample sequence.
  • a batch least squares weight solution is computed for the training data sample sequence to obtain a least square error for each sample.
  • Each error sample is rotated by multiplying by the conjugate of the sample of the reference training sequence and each rotated error sample is numbered (in the order received).
  • a straight line is fit to the imaginary part of the rotated error sample (as a function of sample number): to obtain .a. frequency offset estimate.
  • Each sample in a time slot of the samples of a dataset is multiplied by a complex exponential function of the frequency offset estimate.
  • An estimator for estimating a modulation index and frequency offset of a received continuous-phase-modulated (CPM) signal includes at least two filters for filtering the received CPM signal werein the at least two filters comprise a correlation filter and a low pass filter, a calculator for calculating an ⁇ value and a ⁇ value, and a processor for receiving a signal output by each of the at least two filters, the ⁇ value, and the ⁇ value.
  • the processor is adapted to calculate estimates of the modulation index and frequency offset from the signals received by the processor and the received ⁇ value and ⁇ value.
  • a method of estimating a modulation index and frequency offset of a received continuous-phase-modulated (CPM) signal includes filtering a phase component of a differentially demodulated CPM signal via at least two filters wherein the filtering the phase component comprises filtering the phase component via a correlation filter and a low pass filter calculating an ⁇ value and a ⁇ value, receiving a signal output by each of the at least two filters, the ⁇ value, and the ⁇ value, and calculating estimates of the modulation index and frequency offset from the received signals and the received ⁇ value and ⁇ value.
  • An estimator for estimating a modulation index and frequency offset of a received continuous-phase-modulated (CPM) signal includes a noise whitener for whitening noise of the received CPM signal, at least two filters for filtering the noise-whitened CPM signal, an initializer for processing a training sequence, and a processor for receiving a signal output by each of the at least two filters and the processed training sequence.
  • the processor is adapted to calculate estimates of the modulation index and frequency offset from the signals received by the processor and the processed training sequence.
  • a method of estimating a modulation index and frequency offset of a received continuous-phase-modulated (CPM) signal includes whitening noise of the received CPM signal, filtering the noise-whitened CPM signal via at least two filters, processing a training sequence, receiving a signal output by each of the at least two filters and the processed training sequence, and calculating estimates of the modulation index and frequency offset from the received signals and the processed training sequence.
  • An estimator for estimating a modulation index and frequency offset of a received continuous-phase-modulated (CPM) signal includes at least two filters for filtering the received CPM signal, a noise whitener for whitening noise of a signal output by the at least two filters, an initializer for processing a training sequence, a processor for receiving signals output by the noise whitener and the processed training sequence.
  • the processor is adapted to calculate an estimate of the modulation index and the frequency offset from the received signals and the processed training sequence.
  • ⁇ represents a vector including elements representing scaled versions of estimates of the modulation index and the frequency offset.
  • C represents a noise covariance matrix
  • B represents a data model matrix
  • ⁇ is an observation vector that represents a phase of the CPM signal.
  • estimators associated with aspects of the present invention may be divided into two distinct categories, namely estimators that assume white noise and estimators that assume colored noise. Another classification within the two distinct categories may be made based upon whether or not Inter-Symbol Interference (ISI) is assumed to be present in the input signal. When it is assumed that ISI is present, two further subclasses may be introduced based on whether or not the value of a parameter ⁇ is known or unknown.
  • ISI Inter-Symbol Interference
  • v ( B T C ⁇ 1 B ) ⁇ 1 B T C ⁇ 1 ⁇
  • v is a vector that includes elements representing scaled versions of the estimates of the modulation index and the frequency offset.
  • the matrix C represents a noise covariance matrix and the matrix B represents the data model.
  • the last three terms in equation (1) are a filter operation on an observation vector ⁇ which is the phase input to the estimator.
  • a first approach uses an estimator based upon a simple data model, which does not take into account the Inter-Symbol Interference (ISI).
  • ISI Inter-Symbol Interference
  • the first approach assumes white noise and no ISI.
  • an output i.e., an element of the observation vector
  • b k is a transmitted bit
  • h is a modulation index
  • f is an actual frequency offset
  • T sym is a symbol period
  • n k is a distortion term that includes noise and ISI.
  • the values of x and y of the vector v can be estimated.
  • the estimates of the modulation index h and the frequency offset f may be directly derived from the estimates of x and y respectively.
  • the estimates of the values of x and y of the vector v can be obtained by applying two linear operations on the observation vector ⁇ and a post-processing step that depends on the weight of the training sequence.
  • FIGURE 1 illustrates a block diagram of an estimator 100 in accordance with principles of the present invention.
  • the estimator 100 of FIGURE 1 is based on the data matrix shown in equation (3).
  • the estimator 100 implements the operations of equation (7) and, as mentioned above, assumes white noise and no ISI.
  • a received signal ⁇ k (the signal received, mixed down to base-band, and differentially demodulated) is passed through a first finite-impulse-response (FIR) filter 102 to yield q 1 .
  • the coefficients for the correlation filter 102 are +1 or -1.
  • the received signal ⁇ k is also passed through a second FIR filter 104 to yield q 2 .
  • the training sequence which is a data sequence known at both the receiver and the transmitter, can be used to derive ⁇ and ⁇ .
  • the calculated ⁇ is output to a first multiplier 106A and a fourth multiplier 106D.
  • the derived ⁇ is output to a second multiplier 106B and a third multiplier 106C.
  • q 1 which is output by the first FIR filter 102, is multiplied, at the first multiplier 106A, with the derived ⁇ .
  • q 1 is also multiplied, at the second multiplier 106B, with the derived ⁇ .
  • q 2 which is output by the second FIR filter 104, is multiplied with the derived ⁇ at the third multiplier 106C.
  • q 2 which is output by the second FIR filter 104, is also multiplied with the derived a at the fourth multiplier 106D.
  • the result output by the first multiplier 106A and the result output by the third multiplier 106C are added at a first adder 108A.
  • the result output by the second multiplier 106B and the result output by the fourth multiplier 106D are added at a second adder 108B.
  • the result output by the first adder 108A is x from equation (2). From equation (2), x can be scaled to yield an estimate of the modulation index h. As shown in equation (2), by multiplying x with 1/ ⁇ , the modulation index h is produced. Therefore, at multiplier 110A, x is multiplied with 1/ ⁇ , thereby yielding an estimate of the modulation index h. As is also evident from equation (2), the output of the second adder 108B, y can be multiplied by 1/(2 ⁇ T sym ) at multiplier 110B to produce an estimate of the frequency offset f.
  • the simple estimator 100 might yield biased results. For example, a non-zero mean noise term or correlation between the noise and the desired signal might cause the simple estimator 100 to produce unsatisfactory results.
  • the bias in the estimate of the modulation index typically depends on one or more of the frequency offset, the modulation index, and the value of a signal-to-noise ratio E b / N o . Most typically, no significant bias is present for the estimate of the frequency offset.
  • the bias in the modulation-index estimate can be compensated for at a particular value of E b / N o .
  • the value of E b / N o at which the receiver operates at a bit-error rate (BER) of 10 -3 can be selected.
  • the bias in the modulation-index estimate depends on the modulation index itself, the bias can be compensated for at a typical modulation index value, such as, for example, 0.32.
  • a post-processing step in accordance with principles of the present invention takes into account the fact that the bias in the estimate of the modulation index h depends approximately quadratically on the estimated frequency offset f in order to compensate for the bias in the modulation index estimate.
  • h comp h + c o + c 2 y 2
  • the coefficients c 0 and c 2 are chosen via a curve-fitting process in order to minimize the bias.
  • the bias in the modulation-index estimate and the frequency-offset estimate can be derived by simulation. From the simulation results, adequate bias-reduction processes could be derived via curve fitting.
  • FIGURE 2 illustrates the estimator of FIGURE 1 with additional post-processing to remove bias from the estimate of the modulation index h.
  • the estimator 200 of FIGURE 2 is similar to the estimator 100 of FIGURE 1, except for the implementation of additional components used to introduce the coefficients c 0 and c 2 . As noted above, the coefficients c 0 and c 2 are used to remove bias from the estimate of the modulation index h.
  • y which is output by the second adder 108B, is squared by a squaring block 202.
  • An output of the squaring block 202 is multiplied with the value of c 2 at a multiplier 204.
  • An output of the multiplier 204 is added, at an adder 206, to c 0 and to the estimate of the modulation index h.
  • the estimate of the modulation index h is output by the multiplier 110A.
  • the adder 206 outputs the bias-compensated modulation index h comp .
  • the simple estimators 100 and 200 represent relatively computationally efficient implementations; however, simplification of the data model implemented by the estimators 100 and 200 might not always produce optimal results. Therefore, an estimator based on a more-complex data model than that used in the estimators 100 and 200 can be utilized in another embodiment of the present invention.
  • FIGURE 3 is a block diagram that schematically illustrates a more-complex estimator 300 in accordance with principles of the present invention.
  • the more-complex estimator assumes white noise in a manner similar to that of estimators 100 and 200.
  • the more-complex estimator 300 assumes that ISI is present in the signal and further that the parameter s (see data model from equation (9) shown below) is unknown.
  • the model on which the estimator 300 is based is a more complicated model, namely equation (12) shown below.
  • equation (12) shown below.
  • a relatively simple ISI model has been assumed.
  • other ISI models can be used without departing from principles of the present invention.
  • ⁇ k ⁇ ⁇ k ⁇ 1 + ( 1 ⁇ 2 ⁇ ) ⁇ k + ⁇ ⁇ k + 1
  • Equation (9) shows the relationship between the input phase ⁇ k and the output phase ⁇ k and allows for ISI by the parameter ⁇ . If no ISI is present, then the parameter ⁇ has a value of zero.
  • ⁇ k is the phase of a transmitted symbol a k .
  • Equation (11) shows that the ISI exhibits a relationship with the foregoing bit and the following bit.
  • the estimator 300 which is described mathematically in equations (4), (11), and (12), requires, in addition to the filtering and correlation shown in the estimator 200, another filter, graphically represented as a middle filter 306.
  • the filtering and correlation of filters 302 and 304 operate in a manner similar to filters 102 and 104 of FIGURE 2.
  • the middle filter 306 has N-2 coefficients c k . For the coefficients c k , the following holds: c k ⁇ ⁇ 0, ⁇ 2, ⁇ 4 ⁇ .
  • Variables x and z are manipulated by multipliers 310A and 310B in a manner similar to that shown for x and y in FIGURE 2 to yield the modulation index h and the frequency offset f.
  • the value of the parameter ⁇ from equation (9) may be assumed to be known.
  • the parameter ⁇ is deduced given the overall filter chain in the transmitter and receive parts of the transceiver. Therefore, the estimator 300 may be simplified by assuming, in addition to white noise and ISI, that the parameter ⁇ is known. Due to this fact, the estimator 300 can be simplified and the number of filters utilized reduced as shown in FIGURES 3A and 3B.
  • the estimator 300A includes a correlator filter that is slightly more complex because the filter coefficients are no longer +1 or -1 as in the simple estimators 100, 200, and 300.
  • the implementation of the estimator 300A derived from equation (14) requires two filters: 1) a low-pass filter (304) similar to that in the estimators 100 and 200; and 2) a correlation filter (302) that is matched to the channel (i.e., a matched correlator). Therefore, the N-2 filter coefficients d k are no longer +1 or -1 but take values of the set ⁇ 1, ⁇ (1-2 ⁇ ), ⁇ (1-4 ⁇ ) ⁇ . As such, the correlation filter is more complex than that of the estimators 100, 200, and 300.
  • the modulation index h and the frequency offset f are calculated in a manner similar to that of FIGURE 3.
  • the vector-matrix multiplier 308 outputs variables x and y, which are in turn manipulated by multipliers 310A and 310B to form the estimates of the modulation index h and frequency offset f.
  • the first implementation of the estimator 300A requires the matched correlator.
  • the matched correlator has increased computational complexity; therefore, a second implementation of the estimator 300 with reduced computation complexity, is described below.
  • the second implementation of the estimator of FIGURE 3 is illustrated.
  • the second implementation is not as complex as the first implementation; however, the second implementation is more complex than the estimators 100 or 200.
  • the second implementation includes a post-processing step that needs to be executed only once.
  • a third class of estimators includes noise whitening to further improve the performance of the modulation index h and the frequency offset f estimators.
  • FIGURE 4 is a block diagram of a noise whitening estimator. Due to differential demodulation preceding the estimation, the distortion term n k no longer exhibits typical white noise characteristics. Once the covariance of the matrix of the noise is known, the estimation process can be improved.
  • the noise whitening is performed by multiplication of the inverse C -1 of the noise covariance matrix C.
  • the matrix multiplication of B T C - 1 with the observation vector ⁇ can be implemented in two ways. In a first option, which is used in the estimator of FIGURE 4, this operation is implemented by applying n (n being equal to the number of columns of matrix B) filters in parallel (multiplying with B T ).
  • the operation above is instead performed by two subsequent filter operations, where the first filter operates on the observation vector ⁇ to whiten the noise present in that vector, i.e. by multiplication of C -1 . Then the output of this filter is fed to the n filters in parallel (n being equal to the number of columns in matrix B) , i.e. multiplication by B T . Both options are functionally the same.
  • the whitening of the noise is performed explicitly, while in the first option, the noise whitening is implicitly performed.
  • the estimator described by equation (1) does not restrict the values of the filter coefficients to +1 or -1, thereby increasing complexity over both the estimators 100, 200, and 300.
  • the estimator described by equation (1) is an improved noise-whitening estimator that outperforms the estimators 100, 200, and 300 at the cost of increased complexity.
  • a first option is to quantize the inverse of the noise covariance matrix C. Although the complexity might be reduced, the quantization introduces a performance loss in the estimator.
  • a second option is to adapt the structure of the inverse of the noise covariance matrix C.
  • a finite-impulse-response (FIR) filter may be utilized to whiten the noise. Due to the differential demodulated estimator input signal the noise has a high-pass characteristic. Approximated whitening can be achieved by passing the signal through a low-pass filter.
  • FIR finite-impulse-response
  • K-tap comb filter K may be chosen such that a good balance is obtained between performance loss (compared to ideal whitening) and complexity reduction.
  • a third option would be to implement the approximated whitening operation by means of a low-pass infinite impulse response (IIR) filter.
  • the estimators 100, 200, 300, 300A, and 300B may include colored-noise compensation.
  • the estimators 100, 200, 300, 300A, and 300B may include colored-noise compensation.
  • an estimator similar to that of estimator 100 is obtained, except that the estimator is now colored noise compensated.
  • Substituting the data model from equation (11) into equation (1) yields a noise-whitened estimator similar to that of estimator 300.
  • All of the estimators 100, 200, 300, 300A, and 300B may be altered, by changing the noise model used, to yield colored-noise-compensated estimators.
  • the incoming signal ⁇ is passed to each of a first FIR filter 404 and a second FIR filter 406 in order to be low-pass filtered.
  • the FIR filters 404, 406 implicitly whiten the noise based on the values of the matrix A.
  • An output p of the first FIR filter 404 and an output r of the second FIR filter 406 are similar to the p and r values of the estimator 300B, except for the addition of the noise whitening.
  • the outputs p and r are utilized in further calculations made in a post processor 420.
  • An initialization unit 422 of the estimator 400 receives the training sequence ⁇ b 1 ...b n ⁇ .
  • parameter ⁇ may also be required (see equations (13) and (14)).
  • matrix B which represents the data model
  • matrix B T C - 1 B is calculated.
  • the matrix B T C - 1 B is passed to the post processor 420 as an initialized training sequence and estimates the modulation index h and the frequency offset f in accordance with the above equations.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
  • Stabilization Of Oscillater, Synchronisation, Frequency Synthesizers (AREA)
  • Transmitters (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)

Claims (32)

  1. Schätzmittel (100; 200; 300; 300A; 300B; 400) zum Schätzen eines Modulationsindex (h) und eines Frequenzversatzes (f) von einem empfangenen phasenkontinuierlich modulierten CPM-Signal, dadurch gekennzeichnet, dass das Schätzmittel aufweist:
    zumindest zwei Filter (102, 104; 302, 304; 306; 404, 406) zum Filtern einer Phasenkomponente (Φ) von einem differentiell demodulierten CPM-Signal, wobei die zumindest beiden Filter ein Korrelationsfilter (102; 302; 404) und ein Tiefpassfilter (104; 304; 406) beinhalten;
    ein Berechnungsmittel zum Berechnen eines α-Wertes und eines β-Wertes aus einer Trainingssequenz;
    einen Prozessor (106A, 106B, 106C, 106D, 108A, 108B, 110A, 110B; 308, 310A, 310B) zum Empfangen eines Signals, das durch jedes der zumindest zwei Filter ausgegeben wird, des α-Wertes und des β-Wertes; und
    wobei der Prozessor dazu ausgestaltet ist, um Schätzungen des Modulationsindex und des Frequenzversatzes aus den Signalen, die durch den Mikroprozessor empfangen werden, und dem empfangenen α-Wert und β-Wert zu berechnen.
  2. Schätzmittel nach Anspruch 1, bei dem die zumindest zwei Filter außerdem ein zusätzliches Tiefpassfilter (306) beinhalten.
  3. Schätzmittel nach Anspruch 1, außerdem mit einem Nachprozessor (202, 204, 206) zum Entfernen von Bias aus der Schätzung des Modulationsindex.
  4. Schätzmittel nach Anspruch 3, bei dem der Nachprozessor ausgestaltet ist, um Informationen bezüglich des Frequenzversatzes zu empfangen und um den Modulationsindex zu verarbeiten, um einen kompensierten Modulationsindex (hcomp) zu bilden.
  5. Schätzmittel nach Anspruch 1, bei dem die zumindest zwei Filter FIR-Filter mit begrenztem Ansprechen auf einen Impuls sind.
  6. Schätzmittel nach Anspruch 1, bei dem das Schätzmittel in einer BLUETOOTH-Vorrichtung implementiert ist.
  7. Schätzmittel nach Anspruch 1, außerdem mit:
    einem Mittel zum Rausch-Weißen des empfangenen CPM-Signals.
  8. Schätzmittel nach Anspruch 7, bei dem das Mittel zum Rausch-Weißen konfiguriert ist, um das Rauschen vor den zumindest zwei Filtern zu weißen.
  9. Schätzmittel nach Anspruch 7, bei dem zumindest eines der zumindest zwei Filter das Mittel zum Rausch-Weißen aufweist.
  10. Schätzmittel nach Anspruch 1, außerdem mit:
    einem Mittel zum Rausch-Weißen eines Signal, das von den zumindest zwei Filtern ausgegeben wird.
  11. Schätzmittel nach Anspruch 1, bei dem der Prozessor konfiguriert ist, um abzuschätzen:
    den Modulationsindex und den Frequenzversatz gemäß der vorliegenden Gleichung: ν = ( B T C − 1 B ) − 1 B T C − 1 φ
    Figure imgb0035
    wobei ν einen Vektor darstellt;
    wobei der Vektor Elemente enthält, die skalierte Versionen von Schätzungen des Modulationsindex (h) und des Frequenzversatzes (f) darstellen;
    wobei C eine Rausch-Kovarianz-Matrix darstellt;
    wobei B eine Datenmodell-Matrix darstellt;
    wobei Φ ein Kennvektor ist, der eine Phase von dem CPM-Signal darstellt, und der Kennvektor durch die folgende Gleichung bezeichnet ist: φ k = b k h π + 2 π f T sym + n k
    Figure imgb0036
    wobei bk ein übertragenes Bit darstellt;
    wobei Tsym eine Symbolperiode darstellt; und
    wobei nk einen Rauschterm darstellt.
  12. Schätzmittel nach Anspruch 11, bei dem die Datenmodell-Matrix durch die folgende Gleichung modelliert ist: B = [ b 1 1 b 2 1 b 3 1 ⋮ ⋮ b N 1 ]
    Figure imgb0037
    wobei b1, b2, b3, ...bN Bits von einer Trainingssequenz darstellen.
  13. Schätzmittel nach Anspruch 11, bei dem die Datenmodell-Matrix durch die folgende Gleichung modelliert ist: B = [ b 2 c 2 1 b 3 c 3 1 b 4 c 4 1 ⋮ ⋮ ⋮ b N − 1 c N − 1 1 ]
    Figure imgb0038
    wobei b2, b3, b4, ...bN-1 Bits von einer Trainingssequenz darstellen; und
    wobei C2, C3, C4, ... cN-1 Filterkoeffizienten darstellen.
  14. Schätzmittel nach Anspruch 13, bei dem eine Beziehung zwischen den Bits der Trainingssequenz und den Filterkoeffizienten durch die folgende Gleichung definiert ist: c k = ( b k − 1 − 2 b k + b k + 1 )
    Figure imgb0039
  15. Schätzmittel nach Anspruch 11, bei dem die Datenmodell-Matrix durch die folgende Gleichung modelliert ist: B = [ d 2 1 d 3 1 d 4 1 ⋮ ⋮ d N − 1 1 ]
    Figure imgb0040
    wobei d2, d3, d4, ... dN_1 Filterkoeffizienten darstellen.
  16. Schätzmittel nach Anspruch 15, bei dem eine Beziehung zwischen den Bits der Trainingssequenz und den Filterkoeffizienten durch die folgende Gleichung definiert ist: d k = ( ε b k − 1 + ( 1 − 2 ε ) b k + ε b k + 1 ) ,
    Figure imgb0041
    wobei ε ein Parameter ist, der einen Wert der Inter-Symbol-Interferenz darstellt.
  17. Schätzmittel nach Anspruch 1, bei dem der α-Wert und der β-Wert aus Datenbits (b) der Trainingssequenz und der Anzahl an Bits (N) abgeleitet sind.
  18. Schätzmittel nach Anspruch 1, bei dem der α-Wert durch die folgende Gleichung dargestellt ist: α = N N 2 − S 2 ,
    Figure imgb0042
    wobei S = ∑ k = 1 N b k
    Figure imgb0043
     eine Summierung von allen Datenbits und N die Anzahl an Bits ist.
  19. Schätzmittel nach Anspruch 1, bei dem der β-wert durch die folgende Gleichung dargestellt ist: β = − S N 2 − S 2 ,
    Figure imgb0044
    wobei S = ∑ k = 1 N b k
    Figure imgb0045
     eine Summierung von allen Datenbits und N die Anzahl an Bits ist.
  20. Schätzmittel nach Anspruch 1, bei dem ein x-Wert und ein y-Wert durch Implementieren einer Fehlerquadrat-Technik implementiert sind, und bei dem der Modulationsindex (h) und der Frequenzversatz (f) direkt aus dem x-Wert und dem y-Wert abgeleitet sind.
  21. Schätzmittel nach Anspruch 20, bei dem der Modulationsindex (h) durch Multiplizieren des x-Wertes mit 1/Π abgeleitet ist und der Frequenzversatz (f) durch Multiplizieren des y-Wertes mit 1/(2ΠTsym) abgeleitet ist, wobei Tsym eine Symbolperiode darstellt.
  22. Verfahren zum Schätzen eines Modulationsindex und eines Frequenzversatzes aus einem empfangenen phasenkontinuierlich modulierten CPM-Signals, dadurch gekennzeichnet, dass das Verfahren umfasst:
    Filtern einer Phasenkomponente eines differentiell modulierten CPM-Signals über zumindest zwei Filter, wobei das Filtern der Phasenkomponente das Filtern der Phasenkomponente über ein Korrelationsfilter (102; 302; 404) und ein Tiefpassfilter (104; 304; 406) beinhaltet;
    Berechnen eines α-Wertes und eines β-Wertes aus einer Trainingssequenz;
    Empfangen eines Signals, das durch jeden der zumindest zwei Filter ausgegeben wird, des α-Wertes und des β-Wertes; und
    Berechnen von Schätzungen des Modulationsindex und des Frequenzversatzes aus den empfangenen Signalen und dem empfangenen α-Wert und β-Wert.
  23. Verfahren nach Anspruch 22, bei dem das Filtern der Phasenkomponente eines unterschiedlich demodulierten CPM-Signals über zumindest zwei Filter außerdem das Filtern der Phasenkomponente über ein zusätzliches Tiefpassfilter (306) beinhaltet.
  24. Verfahren nach Anspruch 22, außerdem mit dem Entfernen von Bias aus der Schätzung des Modulationsindex.
  25. Verfahren nach Anspruch 24, bei dem der Schritt des Entfernens von Bias das Empfangen von Informationen bezüglich des Frequenzversatzes und das Manipulieren des Modulationsindex beinhaltet, um einen kompensierten Modulationsindex zu bilden.
  26. Verfahren nach Anspruch 22, bei dem die Schritte in der aufgelisteten Reihenfolge durchgeführt werden.
  27. Verfahren nach Anspruch 22, bei dem die zumindest zwei Filter FIR-Filter mit begrenztem Ansprechen auf einen Impuls sind.
  28. Verfahren nach Anspruch 22, bei dem das Verfahren in einer BLUETOOTH-Vorrichtung implementiert ist.
  29. Verfahren nach Anspruch 22, außerdem mit:
    Weißen des Rauschens des empfangenen CPM-Signals.
  30. Verfahren nach Anspruch 29, bei dem die Schritte in der aufgelisteten Reihenfolge durchgeführt werden.
  31. Verfahren nach Anspruch 29, bei dem der Schritt des Weißens vor dem Schritt des Filterns durchgeführt wird.
  32. Verfahren nach Anspruch 29, bei dem der Schritt des Weißens durch zumindest einen der zumindest zwei Filter durchgeführt wird.
EP03761510A 2002-06-27 2003-06-26 Verfahren und vorrichtung zur gleichzeitigen schätzung des frequenzversatzes und modulationsindex Expired - Lifetime EP1520384B1 (de)

Applications Claiming Priority (5)

Application Number Priority Date Filing Date Title
US603469 1990-10-25
US39211402P 2002-06-27 2002-06-27
US392114P 2002-06-27
US10/603,469 US7277504B2 (en) 2002-06-27 2003-06-25 Method and system for concurrent estimation of frequency offset and modulation index
PCT/EP2003/006721 WO2004004262A1 (en) 2002-06-27 2003-06-26 Method and system for concurrent estimation of frequency offset and modulation index

Publications (2)

Publication Number Publication Date
EP1520384A1 EP1520384A1 (de) 2005-04-06
EP1520384B1 true EP1520384B1 (de) 2006-08-09

Family

ID=32045136

Family Applications (1)

Application Number Title Priority Date Filing Date
EP03761510A Expired - Lifetime EP1520384B1 (de) 2002-06-27 2003-06-26 Verfahren und vorrichtung zur gleichzeitigen schätzung des frequenzversatzes und modulationsindex

Country Status (7)

Country Link
US (1) US7277504B2 (de)
EP (1) EP1520384B1 (de)
CN (1) CN1679291A (de)
AT (1) ATE336130T1 (de)
AU (1) AU2003246589A1 (de)
DE (1) DE60307469T2 (de)
WO (1) WO2004004262A1 (de)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7903769B2 (en) * 2005-12-12 2011-03-08 Telefonaktiebolaget L M Ericsson (Publ) Method and apparatus for phase-noise compensation in digital receivers
US8750444B2 (en) 2011-05-06 2014-06-10 Northrop Grumman Systems Corporation Snapshot processing of timing data
KR101675963B1 (ko) * 2012-09-06 2016-11-15 한국전자통신연구원 무선통신 시스템에서의 변조방식 분류 장치 및 방법
US9306781B2 (en) * 2014-04-01 2016-04-05 Samsung Electronics Co., Ltd Apparatus and method for receiving signal in communication system supporting Gaussian frequency shift keying modulation scheme
US9276704B1 (en) * 2014-12-15 2016-03-01 Intel Corporation Maximum likelihood sequence detection in the phase domain
CN105827551B (zh) * 2015-01-04 2019-02-15 展讯通信（上海）有限公司 一种估计调制指数和频率偏移的方法及系统
US20170180182A1 (en) * 2015-12-22 2017-06-22 Intel Corporation Joint noncoherent demodulation and carrier frequency offset correction based on non-linear filtering
WO2018035761A1 (zh) * 2016-08-24 2018-03-01 深圳天珑无线科技有限公司 一种调制阶数的估算方法和装置
WO2018035760A1 (zh) * 2016-08-24 2018-03-01 深圳天珑无线科技有限公司 一种调制指数的估算方法和装置
CN111063368B (zh) * 2018-10-16 2022-09-27 中国移动通信有限公司研究院 一种音频信号中的噪声估计方法、装置、介质和设备
CN114285710B (zh) * 2021-12-28 2023-09-15 北京升哲科技有限公司 Cpm信号的调制指数估计方法、装置、设备及存储介质
CN114760175B (zh) * 2022-03-21 2023-08-01 上海航天电子通讯设备研究所 基于星载vde的qpsk-cpm分段双向差分解调系统

Family Cites Families (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR2633471B1 (fr) 1988-06-28 1990-10-05 Trt Telecom Radio Electr Procede de demodulation coherente d'un signal module numeriquement en phase continue et a enveloppe constante
US5062123A (en) 1989-08-16 1991-10-29 Cincinnati Electronics Corporation Kalman predictor for providing a relatively noise free indication of the phase of a carrier laden with noise
DE4013384A1 (de) 1990-04-26 1991-10-31 Philips Patentverwaltung Empfaenger mit einer anordnung zur frequenzablagenschaetzung
US5255290A (en) 1992-08-21 1993-10-19 Teknekron Communications System, Inc. Method and apparatus for combined frequency offset and timing offset estimation
WO1994029990A1 (fr) 1993-06-04 1994-12-22 Ntt Mobile Communications Network Inc. Procede de detection de retard par estimation de probabilite maximum et detecteur de retard mettant en ×uvre ce procede
US5802117A (en) 1996-02-08 1998-09-01 Philips Electronics North America Corporation Method and apparatus for joint frequency offset and timing estimation of a multicarrier modulation system
US6275543B1 (en) 1996-10-11 2001-08-14 Arraycomm, Inc. Method for reference signal generation in the presence of frequency offsets in a communications station with spatial processing
AU723679B2 (en) 1997-03-04 2000-08-31 Mitsubishi Denki Kabushiki Kaisha Receiver with a frequency offset correcting function
US6130921A (en) * 1997-12-23 2000-10-10 Motorola, Inc. Frequency shift modulation automatic frequency correction circuit and method
US6363131B1 (en) 1998-04-29 2002-03-26 Hughes Electronics Corporation Method and apparatus for joint timing synchronization and frequency offset estimation
US6389040B1 (en) 1998-09-11 2002-05-14 Lucent Technologies Inc. Apparatus and method for generating a frequency offset estimate for communication systems having frequency selecting fading channels
US6393068B1 (en) 1998-09-22 2002-05-21 Agere Systems Guardian Corp. Communication channel and frequency offset estimator
US6393073B1 (en) 1999-06-28 2002-05-21 Raytheon Company Method of frequency offset estimation and correction for adaptive antennas
KR100396880B1 (ko) * 2000-08-28 2003-09-02 삼성전자주식회사 가변 반송 주파수를 가지는 저잡음 주파수 변조기
US6560303B1 (en) 2000-10-26 2003-05-06 Comsat Corporation Joint frame, carrier and clock synchronization scheme
US6725024B1 (en) 2000-11-07 2004-04-20 Telefonaktiebolaget Lm Ericsson (Publ) Offset local oscillator frequency
CA2363927C (en) 2000-11-28 2004-07-06 Research In Motion Limited Synchronization signal detector and method
US7245672B2 (en) 2002-06-27 2007-07-17 Telefonaktiebolaget Lm Ericsson (Publ) Method and apparatus for phase-domain semi-coherent demodulation

Also Published As

Publication number Publication date
WO2004004262A8 (en) 2004-04-08
DE60307469D1 (de) 2006-09-21
ATE336130T1 (de) 2006-09-15
AU2003246589A1 (en) 2004-01-19
US20040136480A1 (en) 2004-07-15
AU2003246589A8 (en) 2004-01-19
DE60307469T2 (de) 2006-12-28
EP1520384A1 (de) 2005-04-06
WO2004004262A1 (en) 2004-01-08
CN1679291A (zh) 2005-10-05
US7277504B2 (en) 2007-10-02

Similar Documents

Publication Publication Date Title
US7295636B2 (en) Linear single-antenna interference cancellation receiver
Djuric et al. Particle filtering
EP1520384B1 (de) Verfahren und vorrichtung zur gleichzeitigen schätzung des frequenzversatzes und modulationsindex
US8831149B2 (en) Symbol estimation methods and apparatuses
CN101022433B (zh) 高速数字接收机并行自适应盲均衡方法
EP1256215B1 (de) Kleinste quadrate kanalschätzung mit korrektur von farbigem rauschen
CA2448258A1 (en) Channel estimation in cdma communications systems using both lower power pilot channel and higher power data channel
US20040259504A1 (en) Multiuser detection for wireless communications systems in the presence of interference
EP1776814B1 (de) Verfahren zur Berechnung der Filterkoeffizienten für einen Entzerrer in einem Kommunikationsempfänger mittels Hermitischer Optimierung
EP0838111B1 (de) Trägerrückgewinnung für digital phasenmodulierte signale, mittels einer phasemmodulierte bekannten folge
US20040209573A1 (en) DC offset correction in a mobile communication system
EP1183839B1 (de) Verfahren und vorrichtung zur bestimmung der kanal- und baudfrequenzoffsetschätzung mittels einer präambel mit einer wiederholenden sequenz
CN112565132B (zh) 基于混沌成型滤波和相互校验的差分混沌键控方法
EP1190491B1 (de) Verfahren und vorrichtung zur interferenzunterdrückung
US8300746B2 (en) Method and device of frequency domain equalization
US7936813B2 (en) Diversity GMSK-receiver with interference cancellation and methods therein
EP1195032B1 (de) Verfahren zur schätzung und/oder nachführung mit symbolabstand eines schwindenden funkkanals mit fraktionalem abstand
Wu et al. Iterative joint channel estimation and multiuser detection for DS-CDMA in frequency-selective fading channels
EP2304879B1 (de) Ein verfahren und gerät zur demodulation von qam signalen unter verwendung einer symbol-spezifischen amplituden-referenz-schätzung
US6618453B1 (en) Estimating interference in a communication system
US7646832B2 (en) Signal receiver
Míguez et al. A sequential Monte Carlo technique for blind synchronization and detection in frequency-flat Rayleigh fading wireless channels
Vandaele et al. Recursive total least squares algorithm for single-user blind channel equalisation
Chen et al. A sequential Monte Carlo method for blind signal separation using the difference of time delays

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20050113

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL LT LV MK

17Q First examination report despatched

Effective date: 20050621

DAX Request for extension of the european patent (deleted)
GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PT RO SE SI SK TR

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: IT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT;WARNING: LAPSES OF ITALIAN PATENTS WITH EFFECTIVE DATE BEFORE 2007 MAY HAVE OCCURRED AT ANY TIME BEFORE 2007. THE CORRECT EFFECTIVE DATE MAY BE DIFFERENT FROM THE ONE RECORDED.

Effective date: 20060809

Ref country code: BE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: NL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: SI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: AT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: SK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: CZ

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: FI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: RO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: CH

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: LI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

REG Reference to a national code

Ref country code: CH

Ref legal event code: EP

REG Reference to a national code

Ref country code: IE

Ref legal event code: FG4D

REF Corresponds to:

Ref document number: 60307469

Country of ref document: DE

Date of ref document: 20060921

Kind code of ref document: P

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: BG

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20061109

Ref country code: DK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20061109

Ref country code: SE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20061109

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: ES

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20061120

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: PT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070109

NLV1 Nl: lapsed or annulled due to failure to fulfill the requirements of art. 29p and 29m of the patents act
REG Reference to a national code

Ref country code: CH

Ref legal event code: PL

EN Fr: translation not filed
PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed

Effective date: 20070510

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: MC

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20070630

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20070626

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20061110

Ref country code: FR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070511

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: IE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20070626

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20070626

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: EE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: LU

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20070626

Ref country code: CY

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: TR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20060809

Ref country code: HU

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070210

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20220629

Year of fee payment: 20

REG Reference to a national code

Ref country code: DE

Ref legal event code: R071

Ref document number: 60307469

Country of ref document: DE